Best Practices for Agentic Applications

שקד דולצמן, Senior Technical Account Manager, AWS | דורון בלייברג, Generative AI Solutions Architect, AWS | ירדן רון, Engineering Manager, Skai

AWS Summit Tel Aviv 2026 · Building AI and Agentic Applications · 40 min · 16 pages

AI agents are transforming how organizations deliver value, but moving from demo to production demands more than a capable model. This session shares practical best practices to improve agent reliability, reduce latency, cut per-session costs, and make agents maintainable — covering context engineering techniques, multi-agent architectures, tool strategies, and approaches to limit costly agentic loops. Learn how observability, evaluation, and optimization create a continuous improvement flywheel. Hear from Skai on building Celeste AI, an agentic copilot delivering 50% higher operational efficiency and 20% stronger performance across hundreds of brands.

Generated automatically from the recording's transcript. It's meant to replace re-watching, not the recording itself — verify against the video at the timestamp before quoting formally. Slides belong to the speakers and organisers. Produced with recording-to-pdf.